−r ( t 2−t 1)

∂c0 ∂c0
Nos longs, tenho o direito de exercer MAS nos Shorts tenho a obrigação de speculative strategies betting on the direction EUR/USD exchange rate = 1/(EUR/USD)
2
exercer! (Queada)
S0 σ A Currency Thing Dt < K (1−e )
ln +( r + )T
;
= -

= =
Θc ≡ =
1
d1 =
K 2 = d2 +
B Numeraire Price −r ( T −t 2) ∂t ∂T
Dt < K (1−e ) −qT
σ √T rf ∗T
2
−S 0∗N ´ ( d 1 ) σ e
CALL
e
Ende up =1 ->valor que ganhamos no final de invest The Black’s approximation to the price of an American call option consists of: -> computing the prices of
European call options (using the Black-Scholes formula) with expiration at T and any ex-dividend date
e
−rT
-rK

2 √T
. exercise when ST > K, yielding a payoff of ST − K > 0
2 for which early exercise might be optimal

S0 σ Merton´s model q= rf ->picking the highest price of these European calls as the price of the American call ç

+( r− )T
Understanding N(d1) and N(d2)

ln d2=
PUT-CALL PARITY: c0 +k
e
−rT = p +S0
e
−rf ∗T Term N(d2): It is the probability that a call option will be exercised in a risk-neutral
−qT
σ √T
0

K 2 e
world.
N(d2) + qS0N(d1)
Term N(d1): term is not quite so easy to interpret. The expression
Contract size = quantidade de money da moeda
rT
S0 N (d 1 ) e ∂ p0
(Subida)
2
σ √T ln
S0 σ
+( r−Rf + ) T
is the expected stock price at time T in a risk-
Θp ≡ - =

d1=
PUT
=d1- ∂T
neutral world when stock prices less than the strike price are counted as zero. The strike

σ √T K 2 price is only paid if the stock price is greater than K and as just mentioned this has a
probability of N(d2). The expected payoff in a risk-neutral world is therefore
Option Risk and Hedging
−qT
−rT
e N
put option : -S0N(-d1) + K (- d2)
σ √T FI profit ST < K + c0 (short) -> when the FI takes a short call position (sells a call) the FI will
charge c0 + ε and then hedge its short position for a cost c0, generating a small riskless profit of ε −S 0∗N ´ ( d 1 ) σ e −rT
d2=d1-
-> when the FI takes a long call position (buys a call), it will pay c0 − ε and then hedge its long position
e +

σ √T 2 √T
rK
−rT
for a cost −c0, generating again a small riskless profit of ε
Hedging strategies
Protective PUT e
Replicating portfolio VALUE: C0= S0Δ + B0 <=> S0N(d1) -K N(d2) !!ç Risk factor sensitivities

p0= - S0
GBM − Rf∗T −rT −qT
Martigale Xt=E(Xs|Ft), s>t
Volatility
e e N(- d1) +K N(-d2) =x x*N= valor que
e
N(-d2) - qS0N(-d1)

Sem dividendos:

√
¿
Strategy volatility speculation queremos

∂c0 ∂c0
->onde tenho “ coloco o preço tabelado do simbolo *contract size/multiplo *Nº

1 − Rf∗T N(d1) -K −rT N(d2)

∑ ¿¿¿¿ e e
c 0 = S0 contracts (se houver). Se for buscar o valor ao enunciado pequeno, ñ coloco contract
σ= / daily returns: σ= size/múltiplo. Θc ≡ = - =

n−1 Replicate position Ex:(EUR/GBP) (1 contract)

Call atenção ao múltiplo (contract size) no final.
Greeks
Delta Δ
∂t ∂T

√
O que for – é borrow(short) e + é deposit(long) Amount we need to hold the asset to replicate the option
−qT
1 e
−qT N(d1)
e
−rT N(d2) ∂ c 0 −qT ∂ p0 −S 0∗N ´ ( d 1 ) σ e
∑ ¿¿¿¿
∆= B0= -K Dividendos discretos ñ
−rT
∆c ≡
e = *N(d1) ∆p ≡ =
e -rK

n−1 ∂S0 ∂ S0 2 √T
Protective CALL é assim!!
O N(dx) depende se é call ou put
put
Discrete Dividends N(d2)
−qT
S0 =S0* + PV (D) | PV (D) =D
e
−rT ∆= -
e
−qT N(-d1) B0= K
e
−rT N(-d2)
−e *N(-d1)
Δ -> + for cal (increase) / - for put (decrease)
∂ p0
Futures Options When expire ITM, the payoff is: call: (ST -K)(POSITIVE) put (K-ST) (NEGATIVE)
Θp ≡ - =
2
S0 σ Put(short) receive: K -FT,Tf call (long): FT,Tf – K As the expiration date approaches (T goes to 0):
OTM Δ converge to 0 and ∆ converges to 1 (call) or -1 (put) if the option is ITM ∂T
ln +( r + )T Future Price: Ft,Tf=
e
(r −q)(Tf −t )
St ou F0,Tf=
S0 *
GAMA Γ

−qT
d1 =
K 2 = d2 +
(r −q)Tf ∂ C0
2
∂∆c −S 0∗N ´ ( d 1 ) σ e −rT
e e
σ √T
+

2 √T
Γc ≡
2 ❑ rK

∂ S0
<- vai ser usado na formula.Tf é os novos meses (se houver) = =

(r −q)(Tf −t ) * S e ∂ S0
Writing Covered PUT Ft,Tf:
e (r – q - (σ^2)/2) t+ σ
= F0,Tf * e 0 - (σ^2)/2 t+ σ

σ √T
(se pedirem
0
N(-d2)
2
um determinado instante)
−qT −x Θ is linked to the change in the time value of the option In general, Θ is positive(negative) when the

N ´ (d 1) e
time value is negative(positive) when the time value is negative, less time to expiration actually
S0e(r – q - (σ^2)/2) t+ σ onde (r-q) =0 ou seja q=r increases the option price (at expiration the time value is zero) Θ is more negative for ATM options,
2
Mim profit: -p0-c0 when k1≤ St <K2
e
−rT
PUT-CALL PARITY: c0 +k = p0 + S0 – PV(D) Black´s Model is :

−rT N(d1) - K −rT N(d2) −rT N(-d1) - K

= Γp where N’(X) = e onde x
which are those which have more time value: deep-OTM options are worth close to 0 (no IV and almost
no TV), therefore, the passage of time has little impact on the option price deep-ITM options have a

S 0 σ √T
price close to it’s IV (sometimes even lower, when TV <0), since the downside potential is almost as big

e e e
Max: unlimited

√2 π
replicating portfolio: (European Call) C0= F0,Tf | P0= - F0,Tf as the upside potential, hence the TV is close to 0 and the passage of time has little impact on the
option price ->ATM options have lots of TV, since the upside potential is much larger than the downside
potential, hence the passage of time has a big negative impact on the option price ->The TV decreases

−rT faster as the expiration date approaches (mostly for ATM options), unless it is already close to 0 (ITM or

e N(-d2)
= d1(valor)
Γ is nonnegative : -> an increase in S0 makes it more likely that a call will expire ITM, which as we saw
OTM options)

d1=
Rho ρ
increases its ∆ and makes it more likely that a put will expire OTM, which also increases its ∆ (it The ρ measures the sensitivity of the option price to a change in the risk-free interest rate
becomes less negative)
Continuous dividend yield (q)
Merton´s model μ- q 2
Γ is larger for ATM than for ITM or OTM options -> when options are deep-ITM (deep-OTM), they are
almost guaranteed to expire ITM (OTM) -> a change in S0 has little impact on the likelihood of options ∂ c 0 −rT ∂ p0
dSt= (μ-q)*Stdt + σSt dZt | Real world: dSt= (r-q)*Stdt + σSt dZt F 0 , Tf σ expiring ITM vs. OTM, and so it has little impact on ∆ -> by the contrary, when options are ATM, there is
ρc ≡
e = KT *N(d2) ρp ≡ = -KT

ln ⁡( )+(r−r+ )T
a lot of uncertainty about whether they will expire ITM (in which case ∆ should be close to 1 in absolute

Covered CALL
Distribution for St that follows a GBM S0e(r – q - (σ^2)/2) t+ σ value) or OTM (in which case ∆ should be 0) -> therefore, changes in S0 have a large impact on the
∂r ∂r
K 2
likelihood of options expiring ITM vs. OTM and thus on ∆
−qT N(d1) -K −rT N(d2)
e e
As the expiration date approaches: -> Γ decreases if options are ITM or OTM, as it becomes more likely
Black-Scholes call option: c0= S0 that things will remain unchanged and the option will expire the way it is now -> Γ increases if options
−rT
are ATM, since a small change in S0 can be the difference between the option expiring ITM vs. OTM,
e
σ √T
which has a big impact on ∆ *N( -d2)
2
S0 σ Vega(Ѵ)
->ρ is nonnegative for calls and nonpositive for puts ->the higher the risk-free rate, the smaller the

ln +( r−q− )T
present value of the strike paid(received) if the call(put) is exercised ->ρ is higher (in absolute value) for

∂c0
d2= =

( )
2 ITM options the more ITM an option is, the higher the likelihood that it will expire ITM and that the

K 2 F 0 , Tf σ √ T∗N ´ e−qT
strike will be paid/received, the only case where the risk-free rate matters As expiration approaches, we
discount for less time, and so ρ converges to 0

ln +( )T
Mim profit: -2p0-c0 when St=K Vc = Vp ≡ = S0 (d1) where N’(X)

= ∂σ
Dividend-yield rho (ρq)
Max: unlimited
d2=d1-
σ √T K 2 2
The ρq measures the sensitivity of the option price to a change in the dividend-yield

∂c0 ∂ p0
−x −qT *N(d1)
S0e
−qT
σ
2
σ √T 2
ρq,c ≡ = −S0T*
e ρ q,p ≡ = S0T*

LONGE RANGE FORWARD

ln +(r− )T = e onde x = d1(valor) ∂q ∂q
K 2 | d1= d2 σ √T √2 π e
−qT
*N( -d1)
−rT −r∗T
σ √T e e
V is nonnegative: volatility always benefits both calls and puts. V is smaller for ITM and OTM options - -> ρq is nonpositive for calls and nonnegative for puts: ->the higher the dividend yield, the more the
PUT-CALL PARITY: c0 +k = p0 +F0,Tf >than it is for ATM options for deep-OTM options, an increase in volatility will have little impact on the price of the underlying asset tends to go down, which benefits puts and hurts calls -> ρq is close to 0 for
upside potential (i.e. on the likelihood that the option will expire ITM), and no impact on the downside OTM calls, and more negative for ITM calls: ->deep-OTM calls are unlikely to expire ITM, and so an

Price futures-style margining: c0f sm= * c0s sm = F0,Tf* N(d1) -

potential (it can’t get any worse that it already is) -> for deep-ITM options an increase in volatility has a

σ √T
rT increase in the dividend yield and consequent decrease in the drift of the price of the underlying asset

e
positive, but relatively modest, impact on the option price, because it increases both the upside has little impact on the option price by the contrary, ITM calls are likely to be exercised, and thus a
+ potential and the downside potential (the likelihood of the option expiring OTM increases) -> for ATM higher dividend yield has a big impact on option priceÇ
options, an increase in volatility increases the upside potential a lot, with virtually no impact on the
Delta-Gamma-Theta , if Theta tiver no portfolio, como neste exemplo, fazemos da seguinte forma:
K*N(d2) downside potential
−qT −rT
e e
As the expiration date approaches V always decreases ->the closer the expiration date, the less time

p0f sm=
put option : - S0 N(- d1) +K N(-d2)
rT * p s sm = -

e
there is for a larger volatility to materialize in large underlying asset’s price changes
0 F0,Tf* N(-d1) +K*N(- d2) Theta Θ

PUT-CALL PARITY: c0f sm +k = p0f sm +F0,Tf

Risk-neutral probability of expiring ITM: N(d2) for a call OTM: 1-N(d2) or N(- measures the sensitivity of the option price to the passage of time (t), that is, to a decrease in the
time to expiration (T). com dividendos:
d2) | for a put ITM: N(-d2) OTM: N(d2) or 1- N(-d2) 1. Escolher uma option que não as que estão a ser usadas
2ª Parte da Materia “equally likely to expire ITM or OTM?” -> 1º N (d2) = N (−d2) = 0.5 2º d2 ou -d2 = Black´s approximation for American call 2.(Vega) Tini + Thedge= 0 => Tini + NcTc*multiplo => N c=
GEOMETRIC BROWNIAN MOTION( underlying asset)=> dSt= μ*dStdt + σ dSt 0 3º ir à formula do d2 e calculamos o que queremos How to check if it’s optimal to exercise:

dS t PUT-CALL PARITY: c0 +k
e
−rT = p +S −qT ç
0
e
0
If it is optimal to exercise early it will only be the case immediately before an ex-dividend
date and if: 1) the option is deep-ITM at that time 2) and the dividend is large enough.
If the dividend is not large enough, the early exercise just before that ex-dividend date
Tini(valor do enunciado ou ca
dZt ~N(μdStdt; σ2 dS2 tdZt) => = μ*dt + σ dZt ~N(μ dt, σ2 dt)

Tc ( valor da tabela )∗multiplo (s

will never be optimal, no matter how deep ITM the call is. How large?

St Index
Long position worth S0* multiplier
Currency
Distribution for St that follows a GBM: St= S0e(r – (σ^2)/2) t+ σZt = S0e(r – (σ^2)/2) t+ σ 3.Gama hedge -> Δini + Δhedge= Δini + Nc*Δc(price da tabela)*(aqui metemos o múltiplo, se houver)
O nosso contract size costuma ser a 1ª moeda(ativo) do par = Ns (se for positivo , então é – e borrowing. Se for negativo, então é + e deposit)
4. portfolio value: Vini + VT-hedge + VΔ-hedge = V(valor do portfolio) + Nc *price(option tabela)* contract
−rT N(d2)
SRFC
Black-Scholes call option: c0= S0N(d1) -K
e size - Ns* S0
Para ver o valor do portfolio se algo subiu ou desceu: Knock- out options: deixa de existir se tocar na barreira (St= H) durante o tempo da This corresponds to an European Put over the USD/GBP exchange rate with a contract
options R size of 12,000 USD and a strike of 0.77. Since this European Put has the same

1º forma: ∆V =
∂V × ∆σ ou 2º forma: ∆Vnew = Vold+
∂V × (σnew – σold)
Knock- in options: só existe(ativa) quando St atinge a barreira durante o tempo da
options (St = H)
ITM-> above K for call below K for puts
characteristics of the European Call we already priced, we can price the Put via put-call
parity. (Depois do calculos)-> This option can also be a Call(or put), but on the GBP/USD
exchange rate and we did the reverse, so it is impossible to tell if this option is actually

∂σ ∂σ Down-and-in: A opção se torna válida apenas se o preço do ativo cair e atingir a barreira.
Down-and-out: A opção é desativada se o preço do ativo cair e atingir a barreira
Up-and-in: A opção se torna válida apenas se o preço do ativo subir e atingir a barreira.
and Call or Put

2.19
Portfolio greeks Up-and-out: A opção é desativada se o preço do ativo subir e atingir a barreira

dV
∑ wi∗Oi
Produto estruturado

−rT 19

1º e
V= |\ =

dx
1.Dividendos Discretos! 19
*redemption%*Montante = deposit to redemption

−S 2−2800
∑ Wi∗Oi ∑ Wi dOi = * -> risk factor
2º Ru= % * max(

2800
,0 ) =>

dx
R

dx Montante*Ru= Montante*% * max(

R1º
x
−S 2−2800 R:To determine the composition of this portfolio we reverse the procedure we used in a

,0
Delta Hedging depois determinávamos o strike.
previous question to determine the portfolio value and greeks for a portfolio with a
) =>
hedging = ini+ hedge=0 | hedge=
2800
known composition. Since we have only two options (A and B) in this portfolio, we have 2

Δ V V Δ Δ V V Δ N c*

calcular o d2 e N(d2)
depois
unknowns (their positions in the portfolio) and we just need 2 restrictions (we have 4 to
choose from, since we know the portfolio value, Delta, Gamma and Vega). We can use

montante∗%
any combination of restrictions to obtain the portfolio composition with the exception of
ini ini the one that relies on the portfolio Gamma and Vega. This is the case whenever the two

Δ Δ
R
price| if <0 => long (borrow) | > 0 => short (deposit) options have the same maturity, which is the case here. Considering the information
V V about the portfolio value and Delta, we can determine the portfolio positions on options
Gamma-Vega
Exercícios das estratégia Gamma-Vega(exemplo)
2800∗multiplo( se houver ) A (NA contracts) and B (NB contracts) as follows

1. Escolher uma option que não as que estão a ser usadas. * max(2800 -S2,0)*multiplo = Contractos x payoff of 1 long put(or other)(St-K)
2.(Vega) Vini + Vhedge= 0 => Vini + N cV c => N c= contract with K=2800 |2800 -S2= valor de call price or put price -> tabela
3º profit margin for the financial institution:
Vini(valor do enunciado ou calclulado) 1ºvaluing its componentes : π sp= N contract * price (max(2800 -S2,0))*multiplo + (

Vc ( valor da tabela )∗multiplo(se

−rT
e houver )
*redemption*Montante)(valor do 1º)
short 12 contract of A
and long 17 of B
Enunciado Exercicio 3
3.Gama hedge -> Γ ini + Γ hedge= Γ ini + Nc* Γ c(price da tabela)*(aqui metemos o múltiplo, se houver) 3. 17
= Ns Sell price - π sp = profit depois é só

profit
4. COST: Nc* C (price option- tabela)*multiplo (se houver) - Ns*S0 R:The Vega is hedged exactly in the same way as before, with a short position on 4.4363
Delta-Gamma-vega option C contracts, which change the portfolio Delta to -79.1580 EUR. The difference is

= profit margin
To hedge this Delta, we take a short position on 174,618.94 GBP (the underlying asset),
Quando as posições tem a mesma duração 1. that now we will hedge this Delta with a position on futures over the DAX instead of with

{
that is, we borrow 174,618.94 GBP at the risk-free rate of the GBP for 6 months. After all
a position on a portfolio that replicates the DAX (which has a Delta of 1 per “unit”).

TV
hedged sell price e
these hedging positions, the portfolio value becomes:
Before we proceed, we need to determine the Delta of the futures contract over the DAX
for delivery in 2 years.

profit
v hedged
=>

V =¿the par ->

π sp In order to make the portfolio self-financed, and thus hedge the Theta (since the

{
portfolio is already Delta and Gamma hedged), we need to make a deposit of 349,817.94

ini hedge hedge

T V + T p +T c =0
Exercicios:
1. How can you replicate a short position on this option? A LONG é “normal” AUD at the risk-free rate of the AUD for 6 months. R2º: A second, less obvious approach,
interpretação dos valores a short é o inverso, como se fosse um “–“ na formula.
takes advantage of the fact that if all options on the portfolio have the same underlying asset and the
17 22

v ini hedge
+ v hedge
 same maturity (that includes the options in the initial portfolio that needs to be hedged as well as any

V +vp =0 options that are added as hedging positions), then when the portfolio is Gamma neutral it will also be
Vega neutral. We cannot always use this shortcut, because the initial portfolio may already have
c positions in options with different maturities or different underlying assets. But that’s no the case here:

{
all options in the portfolio are over the GBP/AUD exchange rate and have a maturity of 6 months.
2. 21
Therefore, if we use any option over the GBP/AUD with a maturity of 6 months (except, obviously,
ini ❑ ❑ ❑ ❑
T + T ∗N + T ∗N =0 R those that are already on the portfolio) to hedge the Gamma, the Vega will be automatically hedged.
Then we take a position in the GBP to hedge the Delta. And finally, we take a position in the AUD to
V p p c c make the portfolio self-financed and, since it is already Delta and Gamma neutral, also Theta hedged,
thus achieving the desired Delta-Gamma-Vega-Theta hedging

ini ❑ ❑ ❑ ❑ R

v + v ∗N + v ∗N =0
For the second alternative we need to use any option over the GBP/AUD exchange rate with a maturity
of 6 months that is not already in the portfolio in order to hedge the Gamma and Vega simultaneously

V p p c c with this single option position. We don’t have any such option on the table, so we have to make up our
own option. It’s a good idea to use either the put equivalent to the call in the first column of the table,
R or the call equivalent to the put on the second column, since the Gamma and Vega of equivalent calls
(descobrir o Nc e Np) and puts are the same. I will use the put with strike 2.1 and maturity in 6 months (let’s call it p1) and
Exercicio 2 use it to hedge the Gamma
Binariy options (cash-or-nothing & asset-or-nothing)
Payoff:

which means that we need to take a long position on 4.8524 p1 option contracts. This
R position hedges both the Gamma and Vega (you can verify the latter if you want). Next,
we compute the Delta of the portfolio after adding this hedging position. For that, we R Não colocamos theta pq é só 1º ordem
still need to compute the Delta p1

18

Plain-vanilha -> option that have continuos payoff, can generate any payoff in [0, max
payoff) payoff: (european and american options)
European call = long asset-or-nothing call + short cash-or-nothing call (with M = K)
R
aon - con ,M =K
C0
¿ c0 c0 R ouCou D
->To
hedge the Delta we take a short position on 39,434.66 GBP, that is, we borrow 39,434.66
2. 22
European put = long cash-or-nothing put (with M = K) + short asset-or-nothing put GBP at the GBP risk-free interest rate for 6 months. Finally, we determine how much the
portfolio is worth after adding all these hedging positions. To do that, we need first to
con ,M =K - aon
p0
¿p 0 p 0
compute the price of option p1, which we will do using the put-call parity

cash-or-nothing

con = con =
c 0 e−rT M∗N ( d 2) p0
R ->in order to
−rT
e M∗N (−d 2)
achieve Theta neutrality we need to make it self-financed, and so we need to make a
deposit of 80,177.99 AUD at the risk-free rate of the AUD for 6 months. The Delta-
2. 17 Gamma-Vega-Theta hedging is now complete. R:
asset-or-nothing 2. 18

aon = c con = S −qT *N(d1) - K −qT *N(d2) + K

c0 0 +
c0 0
e e
−qT
e *N(d2)

aon con −qT *N(-d2) – [-S −qT *N(-

p0 p0 = - p0 = - K
e 0
e a)
R não usamos 2. 23
contract nem multipler

−qT *N(d2)]
d1) + K
e R: This option gives us the right to do the opposite exchange of currencies the Call gave
Barrier options us. We now have the right to deliver (sell) 12,000 USD, receiving in return 9,240 GBP.
 C e D R:

Pergutas finais
17
R

Exercicio 3
3. 17

R:

19

21
B

18

22

20

R: